CN104866453A - System on a chip, bus interface connection circuit and method for connecting a bus interface - Google Patents

System on a chip, bus interface connection circuit and method for connecting a bus interface Download PDF

Info

Publication number
CN104866453A
CN104866453A CN201510086595.1A CN201510086595A CN104866453A CN 104866453 A CN104866453 A CN 104866453A CN 201510086595 A CN201510086595 A CN 201510086595A CN 104866453 A CN104866453 A CN 104866453A
Authority
CN
China
Prior art keywords
interface
channel
signal
clock
service load
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201510086595.1A
Other languages
Chinese (zh)
Other versions
CN104866453B (en
Inventor
李熙圣
金宇镇
成洛熙
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Samsung Electronics Co Ltd
Original Assignee
Samsung Electronics Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Samsung Electronics Co Ltd filed Critical Samsung Electronics Co Ltd
Priority to CN201911149385.7A priority Critical patent/CN111008170B/en
Publication of CN104866453A publication Critical patent/CN104866453A/en
Application granted granted Critical
Publication of CN104866453B publication Critical patent/CN104866453B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F13/00Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
    • G06F13/38Information transfer, e.g. on bus
    • G06F13/42Bus transfer protocol, e.g. handshake; Synchronisation
    • G06F13/4204Bus transfer protocol, e.g. handshake; Synchronisation on a parallel bus
    • G06F13/4208Bus transfer protocol, e.g. handshake; Synchronisation on a parallel bus being a system bus, e.g. VME bus, Futurebus, Multibus
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F13/00Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
    • G06F13/14Handling requests for interconnection or transfer
    • G06F13/16Handling requests for interconnection or transfer for access to memory bus
    • G06F13/1668Details of memory controller
    • G06F13/1689Synchronisation and timing concerns
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F1/00Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
    • G06F1/04Generating or distributing clock signals or signals derived directly therefrom
    • G06F1/12Synchronisation of different clock signals provided by a plurality of clock generators
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F1/00Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
    • G06F1/26Power supply means, e.g. regulation thereof
    • G06F1/32Means for saving power
    • G06F1/3203Power management, i.e. event-based initiation of a power-saving mode
    • G06F1/3234Power saving characterised by the action undertaken
    • G06F1/324Power saving characterised by the action undertaken by lowering clock frequency
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F13/00Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
    • G06F13/14Handling requests for interconnection or transfer
    • G06F13/36Handling requests for interconnection or transfer for access to common bus or bus system
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F13/00Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
    • G06F13/38Information transfer, e.g. on bus
    • G06F13/40Bus structure
    • G06F13/4004Coupling between buses
    • G06F13/4027Coupling between buses using bus bridges
    • G06F13/405Coupling between buses using bus bridges where the bridge performs a synchronising function
    • G06F13/4059Coupling between buses using bus bridges where the bridge performs a synchronising function where the synchronisation uses buffers, e.g. for speed matching between buses
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F13/00Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
    • G06F13/38Information transfer, e.g. on bus
    • G06F13/40Bus structure
    • G06F13/4063Device-to-bus coupling
    • G06F13/4068Electrical coupling
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F13/00Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
    • G06F13/38Information transfer, e.g. on bus
    • G06F13/42Bus transfer protocol, e.g. handshake; Synchronisation
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02DCLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00Energy efficient computing, e.g. low power processors, power management or thermal management

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Hardware Design (AREA)
  • Information Transfer Systems (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Synchronisation In Digital Transmission Systems (AREA)

Abstract

The present invention provides a system on a chip, a bus interface connection circuit and a method for connecting a bus interface. The system on a chip includes a first interface configured to transmit a payload in synchronization with a first clock signal through a first channel at a first transfer rate; and a second interface that includes: a payload storage connected to the first channel and configured to receive the payload from the first channel; and a payload receiver connected to the payload storage and configured to receive the payload from the payload storage in synchronization with a second clock at a second transfer rate through a second channel. A length of the second channel is shorter than a length of the first channel, and the first clock signal is asynchronous with the second clock signal.

Description

System on Chip/SoC, bus interface connecting circuit and its bus interface method of attachment
The cross reference of related application
This application claims the right of priority of the korean patent application No.10-2014-0019742 submitted in Korean Intellectual Property Office on February 20th, 2014, the open of this application is incorporated herein in full with way of reference.
Technical field
The present invention's design relates to a kind of System on Chip/SoC (SoC), and more particularly, relates to the asynchronous interface in a kind of SoC and method of operating thereof.
Background technology
System on Chip/SoC (SoC) can by a single chip integrated for various complication system.Such as, along with the gathering of the assembly, telecommunication installation, broadcaster etc. of computing machine, special IC (ASIC) or Application Specific Standard Product (ASSP) will more may be realized by SoC.And compact and light and handy mobile device driver is just developed in SoC related industry.
SoC can comprise multiple intellecture property (hereinafter referred to as IP or functional block).Each of multiple functional block can perform their operation.Multiple functional block communicates with one another by the bus in SoC.Such as, Advanced Microcontroller Bus Architecture (AMBA) bus protocol can be used for connecting multiple functional block or IP by AMBA bus.AMBA defines multiple bus layer (or agreement), such as, and Advanced High-Performance Bus (AHB), advanced peripheral bus (APB), Advanced extensible Interface (AXI) etc.AXI bus protocol provides multiple outstanding addressing technique and data cross access technique.
Multiple functional block can communicate with one another in different clock zones or identical clock zone.When functional block operates in identical clock zone, they can synchronously receive and send data.On the other hand, when they operate in different clock zones, functional block can have the additional circuit for receiving and send data asynchronously.
Long distance between transmitter and receiver can cause the long transmission line be connected between transmitter and receiver.In this case, if transmitter and receiver operate in different clock zones, then the output signal of first in first out (FIFO) storer in transmitter is extended when being sent to the receiver for asynchronous interface to send data.The clock signal of receiver is also extended when being sent to transmitter to control the FIFO memory in transmitter.Clock signal synchronization in the output signal of FIFO memory and transmitter, to store data in FIFO memory.But the propagation delay in the output signal of the FIFO memory be extended and the clock signal of receiver can limit the maximum data transfer rate between transmitter and receiver.
Asynchronous bridge or register slice can be used to reduce with the speed compensated between transmitter and receiver as intermediate mass.But these additional circuits can increase power consumption and design complexities.
Summary of the invention
According to the exemplary embodiment of the present invention's design, a kind of System on Chip/SoC can comprise: first interface, and it is configured to send service load by the first channel with the first transfer rate and the first clock signal synchronization; And second interface, it comprises: service load reservoir, and it is connected to the first channel, and is configured to receive service load from the first channel; And service load receiver, it is connected to service load reservoir, and be configured to synchronously receive service load from service load reservoir by second channel with the second transfer rate and second clock, wherein the length of second channel is shorter than the length of the first channel, and the first clock signal and second clock signal asynchronous.
When service load being sent to the second interface from first interface by the first channel, enable signal can be sent to the second interface from first interface.
First interface can comprise local write pointer generator, it is configured to increase local write pointer based on enable signal, and the second interface can comprise remote write pointer generator, it is configured to increase remote write pointer based on the service load received from first interface or enable signal.
First channel can comprise: the first bus line, and it is configured to transmission first clock signal; Second bus line, it is configured to transmit service load; And the 3rd bus line, it is configured to transmit enable signal, and the length of at least one wherein in the first bus line, the second bus line and the 3rd bus line is greater than 2000 μm.
First channel can comprise: the first bus line, and it is configured to transmission first clock signal; Second bus line, it is configured to transmit service load; And the 3rd bus line, it is configured to transmit enable signal, and at least one wherein in the first bus line, the second bus line and the 3rd bus line has at least three impact dampers.
When service load receiver receives service load by second channel from service load reservoir, the second interface transmits the answer signal synchronous with second clock signal by the 3rd channel.
Second interface can comprise local read pointer generator, it is configured to increase local read pointer based on the service load received at service load receiver, and first interface can comprise long-range read pointer generator, it is configured to increase long-range read pointer based on answer signal.
3rd channel can comprise: the 4th bus line, and it is configured to transmit second clock signal; With the 5th bus line, it is configured to transmit answer signal, and the length of at least one wherein in the 4th bus line and the 5th bus line is greater than 2000 μm.
3rd channel can comprise: the 4th bus line, and it is configured to transmit second clock signal; With the 5th bus line, it is configured to transmit answer signal, and at least one wherein in the 4th bus line and the 5th bus line has at least three impact dampers.
According to the present invention design exemplary embodiment, intellecture property (IP) module can comprise: the first channel, its be configured to first clock signal synchronization with first frequency receive service load; Service load reservoir, its be configured to the first clock signal synchronization store service load; Second channel, it is configured to receive service load from service load reservoir; Service load receiver, it is configured to synchronously receive service load by second channel and second clock signal; And the 3rd channel, it is configured to the answer signal synchronous with second clock signal sending second clock signal and receive from service load receiver.
Service load receiver receives the write enable signal of service load by the first channel.According to Advanced extensible Interface (AXI) bus protocol, write enable signal may correspond in writing data valid signal.
According to AXI bus protocol, answer signal may correspond in writing ready for data signal.
According to the exemplary embodiment of the present invention's design, a kind of System on Chip/SoC can comprise: first interface; Second interface; 3rd interface; First channel, it is connected between first interface and the second interface; Second channel, it is connected between first interface and the 3rd interface; And clock channel, it has the first clock signal, second clock signal and the 3rd clock signal, described first clock signal is connected between at least one in first interface and the second interface and the 3rd interface, described second clock signal is connected between first interface and the second interface, and described 3rd clock signal is connected between first interface and the 3rd interface.
System on Chip/SoC also can comprise: the first main IP, and it is connected to first interface; First from IP, and it is connected to the second interface; And second from IP, it is connected to the 3rd interface.
First channel and second channel can according to the communication protocol operations based on AXI bus protocol.
Can by the first data-signal and the first enable signal and the first clock signal synchronization from first interface be sent to the second interface and the 3rd interface at least one.
Second data-signal and the second enable signal and second clock signal synchronously first interface can be sent to from the second interface, or the 3rd data-signal and the 3rd enable signal and the 3rd clock signal synchronization first interface can be sent to from the 3rd interface.
Can by the first answer signal and the first clock signal synchronization from first interface be sent to the second interface and the 3rd interface at least one.
Second answer signal and second clock signal synchronously first interface can be sent to from the second interface, or the 3rd answer signal and the 3rd clock signal synchronization first interface can be sent to from the 3rd interface.
First interface can comprise local write pointer generator, and at least one in the second interface and the 3rd interface comprises remote write pointer generator.
At least one in second interface and the 3rd interface can comprise local read pointer generator, and first interface can comprise long-range read pointer generator.
When the frequency of the first clock signal is equal to or greater than 500Mhz, the length of the first channel can be greater than 2000 μm.
According to the exemplary embodiment of the present invention's design, a kind of bus interface method of attachment of System on Chip/SoC can comprise step: with the first transfer rate, service load and the first clock signal synchronization are sent to buffer memory the second interface by the first channel from first interface; And with the second transfer rate, service load is sent to service load receiver from buffer memory by the second channel asynchronous with the first channel, wherein the length of the first channel is longer than the length of second channel.
Buffer memory can be first in first out (FIFO) storer.
The bus interface method of attachment of described System on Chip/SoC also can comprise step: with the first transfer rate, the enable signal of service load and the first clock signal synchronization are sent to the second interface from first interface by the first channel.
The bus interface method of attachment of described System on Chip/SoC also can comprise step: with the second transfer rate, second clock signal and the answer signal synchronous with second clock signal are sent to first interface from the second interface by the 3rd channel.
Bus interface can be performed based on AXI bus protocol to connect.
When the frequency of the first clock signal is equal to or greater than 500Mhz, the length of the first channel can be greater than 2000 μm.
According to the exemplary embodiment of the present invention's design, a kind of bus interface connecting circuit can comprise: sender interface, and it is configured to send service load, write enable signal and transmitter clock signal by the first channel; And receiver interface, it comprises: FIFO memory, and its remote write pointer be configured to based on being produced based on write enable signal by receiver interface stores service load; And service load receiver, it is configured to read service load from FIFO memory, and wherein, receiver interface is by second channel transmitter-receiver clock and answer signal, and the length of second channel corresponds to the length of the first channel.
Service load can be sent to sender interface from memory assembly.
Can be connected to Memory Controller by service load receiver, described Memory Controller is configured to control store assembly.
Service load and transmitter clock signal synchronously can be latched and be sent to receiver interface.
Service load and transmitter clock signal synchronously can be stored in FIFO memory.
According to the exemplary embodiment of the present invention's design, a kind of asynchronous interface comprises: first interface, and it is configured to transmission and writes data, write enable signal and transmitter clock, and second interface, it is configured to be received by the first bus line write data, receive write enable signal by the second bus line and pass through the 3rd bus line receiver transmitter clock, wherein, first bus line is included in long hop channel to the 3rd bus line, second interface is also configured to the transmitter clock being produced recovery by the adjustment clock skew write between data and transmitter clock, data will be write based on the transmitter clock recovered to store in memory, and data will be write by the first channel shorter than long hop channel and be sent to receiver asynchronously from storer.
First interface is also configured to read answer signal by long hop channel from the second interface receiver clock and first, the receiver clock of recovery is produced by adjusting the first clock skew read between answer signal and receiver clock, read answer signal based on first to produce and synchronous with the receiver clock of recovery second read answer signal, read answer signal based on synchronous with the receiver clock recovered second to produce and read increment signal, and increment signal will be read by the second channel shorter than long hop channel and be sent to transmitter asynchronously.
Accompanying drawing explanation
Describe the exemplary embodiment of the present invention's design in detail by referring to accompanying drawing, the above and other feature of the present invention's design will become clearly, wherein:
Fig. 1 is the block diagram of System on Chip/SoC (SoC) of the exemplary embodiment according to the present invention's design;
Fig. 2 is the block diagram of asynchronous interface circuit of the exemplary embodiment according to the present invention's design;
Fig. 3 is the sequential chart of the operation illustrated according to the service load forwarder in Fig. 2 of the exemplary embodiment of the present invention's design;
Fig. 4 is the sequential chart of the operation illustrated according to the response forwarder in Fig. 2 of the exemplary embodiment of the present invention's design;
Fig. 5 is the block diagram of asynchronous interface circuit of the exemplary embodiment according to the present invention's design;
Fig. 6 is the block diagram of SoC of the exemplary embodiment according to the present invention's design;
Fig. 7 is the process flow diagram of the method illustrated according to the first asynchronous interface in the application drawing 2 of the exemplary embodiment of the present invention's design;
Fig. 8 is the process flow diagram of the method illustrated according to the second asynchronous interface in the application drawing 2 of the exemplary embodiment of the present invention's design;
Fig. 9 is the block diagram of SoC of the exemplary embodiment according to the present invention's design;
Figure 10 is the block diagram of memory interleaving device of the exemplary embodiment according to the present invention's design;
Figure 11 is the block diagram of Advanced extensible Interface (MAXI) channel of multiple amendments of exemplary embodiment according to the present invention's design;
Figure 12 illustrates according to the main interface in the SoC of the exemplary embodiment of the present invention's design and the block diagram from the affairs between interface;
Figure 13 is the block diagram of SoC of the exemplary embodiment according to the present invention's design;
Figure 14 is the block diagram of the data handling system of the SoC of the exemplary embodiment comprised according to the present invention's design;
Figure 15 is the block diagram of the data handling system of the SoC of the exemplary embodiment comprised according to the present invention's design;
Figure 16 is the block diagram of the data handling system of the SoC of the exemplary embodiment comprised according to the present invention's design; And
Figure 17 is the block diagram of the computer system of the SoC of the exemplary embodiment comprised according to the present invention's design.
Embodiment
Hereinafter, each exemplary embodiment of the present invention's design is more fully described with reference to the accompanying drawings.But the present invention's design can realize in many different forms, and should not be construed the embodiment being limited to and setting forth herein.In the accompanying drawings, for the sake of clarity, the size in layer and region and relative size can be exaggerated.Identical Reference numeral refers to identical element in the specification and illustrated in the drawings all the time.
Fig. 1 is the block diagram of System on Chip/SoC (SoC) of the exemplary embodiment according to the present invention's design.Transmitter circuit 100 as the functional block in SoC 150 and acceptor circuit 200 is comprised with reference to Fig. 1, SoC 150.Here, transmitter circuit 100 and acceptor circuit 200 send or receive data, such as, and service load.
Transmitter circuit 100 can comprise as the IP 160 from intellecture property (IP) and the first interface 120 as asynchronous main interface, and acceptor circuit 200 can comprise as the 2nd IP 260 of main IP with as asynchronous the second interface 220 from interface.On the other hand, one IP 160 of transmitter circuit 100 can be main IP, and the first interface 120 of transmitter circuit 100 can be asynchronous from interface, and the 2nd IP 260 of acceptor circuit 200 can be from IP, and the second interface 220 of acceptor circuit 200 can be asynchronous main interface.
Although can be defined as from IP and main IP respectively by an IP 160 and the 2nd IP 260, an IP 160 and the 2nd IP 260 is defined as main IP and from IP by direction respectively that can transmit according to data.In other words, in FIG, an IP 160 can be defined as the producer providing data, and the 2nd IP 260 can be defined as the consumer receiving data.
One or more channel can be connected between first interface 120 and the second interface 220.Conveniently explain, below describe and will concentrate in Fig. 1 two channel B10 and B20 be connected between first interface 120 and the second interface 220.
The first channel B10 between first interface 120 and the second interface 220 can be included in the multiple bus lines sending service load, write enable signal and transmitter clock signal between first interface 120 and the second interface 220.First channel B10 also can be included between first interface 120 and the second interface 220 and send the multiple bus lines reading answer signal and receiver clock signal.
The second channel B20 similar to the first channel B10 can be included in and send service load, write enable signal between first interface 120 and the second interface 220, read multiple bus lines of answer signal, transmitter clock signal and receiver clock signal.
Long length of jumping (long-hop) channel LH for the first channel B10 between the first interface 120 in SoC 150 and the second interface 220 and second channel B20 can increase along with the size of SoC150 and increase.Be inserted in asynchronous bridge in the middle of transmission line or register slice can avoid the sequential of clock signal or service load in violation of rules and regulations, but due to additional circuit thus can increasing circuit complexity and power consumption.In other words, the less circuit for the transmission by long channel (hereinafter referred to as long hop channel) between first interface 120 and the second interface 220 can increasing power efficiency reduce design complexities.
The route length of long hop channel can be defined based on the operating frequency of the signal transmitted between transmitter circuit 100 and acceptor circuit 200.Such as, when operating frequency is greater than 500Mhz, the Minimal routing length of long hop channel may be defined as and is greater than 2000 μm.Operating frequency is larger, and the Minimal routing length of long hop channel can be less.In addition, the route length of the long hop channel in SoC 150 can be greater than 1/2 or 2/3 of the length of the longer direction in the vertical direction of the chip size of SoC 150 or horizontal direction.In some cases, when the operating frequency between transmitter circuit 100 and acceptor circuit 200 is more than 1GHz, the route length of long hop channel can be greater than 500 μm.In addition, the relation between the operating frequency of length, long hop channel and route length can change according to the semiconductor technology feature of such as low-power and high-performance and so on.
According to the exemplary embodiment of the present invention's design in Fig. 2, compared with having the asynchronous circuit of extra intermediate circuit in the long channel in SoC, asynchronous interface can improve performance and reduce power consumption simultaneously.
Fig. 2 is the block diagram of asynchronous interface circuit 105 of the exemplary embodiment according to the present invention's design.See figures.1.and.2, asynchronous interface circuit 105 comprises first interface 120 and the second interface 220.
Conveniently explain, below describe and will concentrate on the long hop channel LH formed by the first channel B10 between the first interface 120 in Fig. 1 and the second interface 220.The channel B10 as long hop channel L20 in Fig. 2 may correspond to the first channel B10 as long hop channel LH in Fig. 1.
The first bus line B11, the second bus line B12, the 3rd bus line B13, the 4th bus line B21 and the 5th bus line B22 can be comprised with reference to Fig. 2, the first channel B10.
With reference to Fig. 2, input effective load data (IPD) signal and effective (IPV) signal of input service load can be inputed to first interface 120.First interface 120 exportable output service load ready (OPR) signal.Effective load data can be inputed to the transmitter 130 in first interface 120 by IPD signal.IPV signal can be the useful signal of the validity of instruction IPD signal.OPR signal can be instruction second interface 220 for receiving the ready signal of the ready state of effective load data.
Second interface 220 exportable output effective load data (OPD) signal and output service load effectively (OPV) signal.Input service load ready (IPR) signal can be inputed to the service load receiver 240 in the second interface 220.OPD signal can be output effective load data.OPV signal can be the useful signal of the validity of instruction OPD signal.IPR signal can be the ready signal that instruction service load receiver 240 is in the ready state for receiving OPD signal.
The first interface 120 operated as the interface for data producer can comprise transmitter 130 and receiver 140.The second interface 220 operated as the interface for data consumer can comprise service load reservoir 230 and service load receiver 240.
Transmitter 130 in first interface 120 can comprise service load input block 2, trigger 4, gating unit 6, export control signal generator 8, trigger 10, local write pointer generator 12, comparer 14, synchronizer 16 and transmitter clock generator 18.Service load can be sent to the second interface 220 by transmitter 130.
The effective load data received can be sent to trigger 4 from IPD Signal reception effective load data F3 by service load input block 2.
Trigger 4 can store the effective load data F3 received from service load input block 2.The effective load data of the exportable latch of trigger 4 is as writing data output signal (O_WDATA) F6.Second interface 220 by first bus line B11 receive from first interface 120 write data output signal (O_WDATA) F6 obtain write data input signal (I_WDATA) F8.
Gating unit 6 can produce writes indicator signal (WPTR_IND) F2, and its instruction have input effective load data when both IPV signal and OPR signal activate from IPD signal.Gating unit 6 can be realized according to AND logic gate.
Export control signal generator 8 and can produce increment signal (WPTR_INC) to increase local write pointer F4.
The exportable local write pointer F4 produced in transmitter 130 of local write pointer generator 12, and increase local write pointer F4 when increment signal (WPTR_INC) activates.
Trigger 10 can latch increment signal and the increment signal of output latch writes enable output signal (O_WEN) F5 as the service load reservoir 230 be sent to by the second bus line B12 in the second interface 220.
Synchronizer 16 can latch and output signal (O_TCLK) F1 synchronously from the long-range read pointer G8 that receiver 140 receives with the transmitter clock in transmitter 130, and exports through synchronous long-range read pointer F14.Synchronizer 16 can be realized according to one group of latch or trigger.
Local write pointer F4 can compare with through synchronous long-range read pointer F14 by comparer 14, and produces OPR signal.Whether OPR signal can indicate FIFO memory 34 that is in the second interface 220 and that be positioned at away from first interface 120 full.
Transmitter clock generator 18 can produce transmitter clock output signal (O_TCLK) F1 being sent to the second interface 220 from first interface 120 by the 3rd bus line B13.There is provided transmitter clock to output signal (O_TCLK) F1 by the IP 160 in Fig. 1, and transmitter clock can be outputed signal (O_TCLK) F1 by transmitter clock generator 18 is passed to the second interface 220.
Receiver 140 in first interface 120 can comprise trigger 20, enable signal generator 22, long-range read pointer generator 24 and clock recovery unit 26.
Clock recovery unit 26 can receive receiver clock input signal (I_RCLK) G6 from the second interface 220 by the 5th bus line B22, and produces the receiver clock RCLK1 recovered.Clock recovery unit 26 exportable receiver clock input signal (I_RCLK) G6 does not carry out any amendment as the receiver clock RCLK1 recovered, or utilizes the restoring circuit of such as delay phase-locked loop (DLL) or delay buffer and so on to export the receiver clock RCLK1 recovered.Clock recovery unit 26 can for the clock synchronous read between response input signal (I_RACK) G5 and the receiver clock signal RCLK1 of recovery to control clock skew.Therefore, when long hop channel L20 is connected between first interface 120 and the second interface 220, the receiver 140 in first interface 120 can operate in the clock zone identical with the clock zone of the service load receiver 240 in the second interface 220.
Trigger 20 can receive from the second interface 220 by the 4th bus line B21 be sent to first interface 120 read reply input signal (I_RACK) G5.Trigger 20 can based on read reply input signal (I_RACK) G5 produce with recovery receiver clock signal RCLK1 synchronous read answer signal G7.
Enable signal generator 22 can produce read increment signal G7B based on the answer signal G7 that reads synchronous with the receiver clock signal RCLK1 recovered.
Long-range read pointer generator 24 can produce the long-range read pointer G8 produced in first interface 120, and when reading to increase long-range read pointer G8 when answer signal G7 activates.Long-range read pointer generator 24 can comprise the multiple triggers for storing long-range read pointer G8.
Service load reservoir 230 can comprise trigger 30, selector switch 32, FIFO memory 34, trigger 36, clock recovery unit 38 and remote write pointer generator 42.
Trigger 30 can be comprised for storing the trigger writing data F11 writing data input signal (I_WDATA) F8 output latch received from the transmitter 130 first interface 120 by the first bus line B11.
Selector switch 32 can based in the second interface 220 produce remote write pointer F12 select FIFO memory 34 target entries with write latch write data F11.Selector switch 32 can be realized according to multiple logic gate, select signal to select the entrance of FIFO memory 34 to produce.
Trigger 36 can latch writes enable input signal (I_WEN) F7, and writing enable input signal (I_WEN) F7 is from the signal writing the delay that enable output signal (O_WEN) F5 obtains by the second bus line B12.Trigger 36 can produce write enable signal (WEN_DST) F10 of the latch synchronous with the transmitter clock signal TCLK1 recovered, and the transmitter clock signal TCLK1 of recovery produces in the second interface 220.
When write enable signal (WEN_DST) F10 latched activates, the data F11 that writes latched can be stored to the entrance selected by selector switch 32 by FIFO memory 34.
Clock recovery unit 38 by the 3rd bus line B13 receiver transmitter clock input signal (I_TCLK) F9, and produces the transmitter clock signal TCLK1 recovered.Clock recovery unit 38 exportable transmitter clock input signal (I_TCLK) F9 does not carry out any amendment as the transmitter clock signal TCLK1 recovered, or exports the transmitter clock signal TCLK1 utilizing the restoring circuit of such as DLL or delay buffer and so on to recover.Clock recovery unit 38 can control clock skew for the clock synchronous between the transmitter clock signal TCLK1 writing data input signal (I_WDATA) F8 and recovery.Therefore, when long hop channel L20 is connected between first interface 120 and the second interface 220, the service load reservoir 230 in the second interface 220 can operate in the clock zone identical with the clock zone of the transmitter 130 in first interface 120.
Remote write pointer generator 42 can produce the remote write pointer F12 produced in the second interface 220, and increases remote write pointer F12 when write enable signal (WEN_DST) F10 latched activates.Remote write pointer generator 42 can comprise the multiple triggers for storing remote write pointer F12.
Remote write pointer generator 42 can comprise logic gate 40, to export the remote write pointer F12 of increase.Remote write pointer generator 42 can latch the remote write pointer F12 of the increase synchronous with the transmitter clock signal TCLK1 recovered.
Service load receiver 240 in second interface 220 can comprise multiplexer 37, gating unit 50, export control signal generator 52, trigger 54, synchronizer 56, comparer 58, trigger 60 and receiver clock generator 62.
Multiplexer 37 can select the entrance of FIFO memory 34 based on local read pointer G3.
When both IPR signal and OPV signal all activate, gating unit 50 can produce indicator signal G2A, and it indicates the entrance of service load receiver 240 according to the selection of local read pointer G3 FIFO memory 34 by OPD signal-obtaining.Gating unit 50 can be realized according to AND logic gate.
Export control signal generator 52 and can produce increment signal G2 to increase local read pointer G3.
Synchronizer 56 can output signal (O_RCLK) G1 and synchronously latch the remote write pointer F12 received from service load reservoir 230 with the service load receiver clock in service load receiver 240, and exports synchronous remote write pointer G3B.Synchronizer 56 can be realized according to one group of latch or trigger.
Local read pointer G3 can compare with synchronous remote write pointer G3B by comparer 58, and produces OPV signal.Whether OPV signal can indicate FIFO memory 34 full.The 2nd IP 260 in Fig. 1 can determine whether to read FIFO memory 34 according to OPV signal.
Trigger 60 can latch increment signal G2 and read response output signal (O_RACK) G4 by the increment signal of the 4th bus line B21 output latch as the receiver 140 be sent in first interface 120.
Receiver clock generator 62 can produce service load receiver clock output signal (O_RCLK) G1 being sent to first interface 120 from the second interface 220 by the 5th bus line B22.There is provided effective load-receiver clock output signal (O_RCLK) G1 by the 2nd IP 260 in Fig. 1, and receiver clock generator 62 can transmit service load receiver clock output signal G1.
In fig. 2, by the first bus line B11 as the long transmission line in long hop channel L20 the IPD signal in first interface 120 is sent to the FIFO memory 34 in the second interface 220.Short channel L10 between multiplexer 37 with FIFO memory 34 can be very short compared with long hop channel L20, thus make service load receiver 240 and service load reservoir 230 can be arranged as close to each other in the layout of SoC 150.On the other hand, first interface 120 and the second interface 220 can be arranged as away from each other in the layout of SoC 150.
By the second bus line B12 as the long transmission line in long hop channel L20, enable output signal (O_WEN) F5 that writes in first interface 120 is sent to the second interface 220.
By the 3rd bus line B13 as the long transmission line in long hop channel L20, transmitter clock output signal (O_TCLK) F1 in first interface 120 is sent to the second interface 220.
IPD signal is postponed by multiple impact damper Bu1, Bu2 and Bu3 on the first bus line B11.Write enable signal F5 is postponed by multiple impact damper Bu10, Bu20 and Bu30 on the second bus line B12.By multiple impact damper Bu11, Bu21 and Bu31 delay pick-off clock output signal (O_TCLK) F1 on the 3rd bus line B13.
When short channel L10 is connected between service load reservoir 230 and service load receiver 240, the service load reservoir 230 in the second interface 220 can operate in the clock zone different from the clock zone of the service load receiver 240 in the second interface 220.
When short channel L11 is connected between receiver 140 and transmitter 130, the receiver 140 in first interface 120 can operate in the clock zone different from the clock zone of the transmitter 130 in first interface 120.
If performed the transmission of service load by the first bus line B11 with the first transfer rate, then can pass through the transmission that short channel L10 performs service load by the second transfer rate.Here, the second transfer rate can be greater than or less than the first transfer rate.
When performing the transmission of service load with the first transfer rate by the first bus line B11, the first transfer rate transmission enable output signal (O_WEN) F5 and transmitter clock output signal (O_TCLK) F1 can be write.
On the other hand, when performing the transmission of service load with the first transfer rate by the first bus line B11, the second transfer rate transmission response output signal (O_RACK) G4 and service load receiver clock output signal (O_RCLK) G1 can be read.Here, the second transfer rate can be greater than or less than the first transfer rate.
A kind of method that asynchronous bus interface connects can comprise step: service load be sent to as the FIFO memory 34 the second interface 220 of main interface using the first transfer rate from as the first interface 120 from interface by the first long hop channel L20.The method that asynchronous bus interface connects also can comprise step: by the length first short channel L10 shorter than the length of the first long hop channel L20 service load to be sent to the service load receiver 240 the second interface 220 from FIFO memory 34 with the second transfer rate.
According to the exemplary embodiment of the present invention's design described in Fig. 2, FIFO memory 34 is arranged in the second interface 220 of transmitter 130 distance in first interface 120 with the multiplexer 37 being connected to FIFO memory 34.Therefore, when the IP 160 in Fig. 1 and the 2nd IP 260 operates in different clock zones, one IP 160 and/or the 2nd IP260 synchronously can send service load with that transmitter clock produced sending service load from an IP 160 and the 2nd IP 260 in long channel (such as, long hop channel).Here, the answer signal corresponding with service load is synchronously sent by long hop channel with receiver clock.Therefore, according to the exemplary embodiment of the present invention's design, the speed that can reduce the circuit between an IP 160 and the 2nd IP 260 limits and complexity.
Can realize each of first interface 120 and the second interface 130 in IP, described IP can be the soft IP module described by hardware description language (HDL).The description of HDL can have the different levels of such as behavior arrangement, register level and transistor level.Soft IP module can be included in the design library that can be provided for Top-Down by Foundry Inc. (foundry company) or IP company.
The IP operated as main IP can be CPU (central processing unit) (CPU), direct memory access (DMA), Graphics Processing Unit (GPU), Video Codec, digital signal processor (DSP), image-signal processor (ISP) and display controller, and described display controller supports the multiple display related port of such as RGB (RGB), high-definition media interface (HDMI), display port, TV (TV) output etc.Dynamic RAM (DRAM) Memory Controller, static RAM (SRAM) Memory Controller and multiple IP special function register (SFR) and the such as peripheral hardware of sound interface (I2S), Sony/philips digital interface form (SPDIF) and so between universal asynchronous receiver/transmitter (UART), internal integrated circuit (I2C), integrated chip is can be as the IP operated from IP.
Writing data output signal (O_WDATA) F6, write enable output signal F5 and read to reply output signal (O_RACK) G4 each transmit during, each of trigger 4,10 and 60 can be used for reducing skew.Here, the skew of the one-period being less than transmitter clock output signal (O_TCLK) F1 can be kept.
According to the exemplary embodiment of the present invention's design described in Fig. 2, at least three impact dampers at least one had bus line B11, B12, B13, B21 and B22 in bus line B11, B12, B13, B21 and B22.But the quantity of the impact damper on bus line B11, B12, B13, B21 and B22 can change according to the length of bus line B11, B12, B13, B21 and B22 or electric capacity.Such as, the quantity of the impact damper on bus line B11, B12, B13, B21 and B22 can be greater than three.
The present invention conceives the signal and assembly that are not limited in Fig. 2.
Fig. 3 is the sequential chart according to the service load transfer operation in Fig. 2 of the exemplary embodiment of the present invention's design.
Reference Fig. 3, F1 are that the transmitter clock produced by the transmitter clock generator 18 in Fig. 2 outputs signal O_TCLK.F2 be produced by the gating unit 6 in Fig. 2 write indicator signal (WPTR_IND).F3 is the service load IPD as the input data in the service load input block 2 inputed in Fig. 2.F4 is the local write pointer signal (WPTR_LCL) produced by the local read pointer generator 12 in Fig. 2.F5 is the write enable signal (O_WEN) produced by the trigger 10 in Fig. 2.F6 is the effective load data writing the latch of data output signal (O_WDATA) as exporting from the trigger 4 in Fig. 2.
F7 writes enable input signal (I_WEN) in the second interface 220, and it is the delay version writing enable output signal (O_WEN) F5 in first interface 120.Enable output signal (O_WEN) F5 that writes in first interface 120 is propagated by impact damper Bu10, Bu20, the Bu30 on the second bus line B12 in long hop channel B10, and the trigger 36 be input in the second interface 220, as writing enable input signal (I_WEN) F7.
F8 writes data input signal (I_WDATA) in the second interface 220, and it is the delay version writing data output signal (O_WDATA) F6 in first interface 120.Data output signal (O_WDATA) F6 that writes in first interface 120 is propagated by impact damper Bu1, Bu2, the Bu3 on the first bus line B11 in long hop channel B10, and the trigger 30 be input in the second interface 220, as writing data input signal (I_WDATA) F8.
F9 is the transmitter clock input signal (I_TCLK) in the second interface 220, and it is the delay version of transmitter clock output signal (O_TCLK) F1 in first interface 120.Transmitter clock output signal (O_TCLK) F1 in first interface 120 is propagated by impact damper Bu11, Bu21, the Bu31 on the 3rd bus line B13 in long hop channel B10, and the clock recovery unit 38 be input in the second interface 220, as transmitter clock input signal (I_TCLK) F9.
F10 is and the transmitter clock signal TCLK1 of the recovery produced in the second interface 220 synchronously write enable signal (WEN_DST) of latch that exports of slave flipflop 36.
F11 be with the transmitter clock signal TCLK1 of the recovery produced in the second interface 220 synchronously the latch that exports of slave flipflop 30 write data (WDATA_DST).
F12 is the remote write signal-arm (WPTR_RMT) produced in service load reservoir 230 based on the write enable signal received (I_WEN).Although first interface 120 is separated with distance L20 by long hop channel B10 with the second interface 220, remote write signal-arm (WPTR_RMT) F12 can make the second interface 220 follow local write pointer (WPTR_LCL) F4 in first interface 120.
F13 can be the output signal of the effective load data (OPD) be stored in each entrance of FIFO memory 34.
D0-D5 corresponds to data, and the 0-6 of F4 corresponds to local write pointer value, and the 0-6 of F12 corresponds to remote write pointer value.
Transmitter clock can be outputed signal (O_TCLK) F1 and be provided to trigger 4 and 10 in the transmitter 130 of the first interface 120 in Fig. 2 and local read pointer generator 12.Transmitter clock input signal (I_TCLK) F9 is provided to the trigger 30 and 36 in the service load reservoir 230 of the second interface 220 in Fig. 2.If transmitter clock input signal F9 had controlled clock skew before being provided to trigger 30 and 36, then recover transmitter clock input signal F9 by clock recovery unit 38.
Signal F1, F5 and F6 can start from first interface 120, and are delayed by such as more than a clock period, as shown in the arrow A R1 in Fig. 3.Delay shown in arrow A R1 can be due to the layout routing delay based on temperature or route conditions, and described route conditions is such as the length of the route conductors of the layout route in SoC 150, width, resistance or stray capacitance.
When first interface 120 is bide one's time before transmitter 130 can transmit service load etc., transmitter clock output signal (O_TCLK) F1 can stop triggering; In addition, transmitter clock output signal (O_TCLK) F1 can always trigger.Second interface 220 by recovering transmitter clock output signal (O_TCLK) F1 with under type, that is, produces the transmitter clock TCLK1 recovered from transmitter clock input signal (I_TCLK) F9 through the 3rd bus line B13.
Fig. 4 is the sequential chart according to the response transfer operation in Fig. 2 of the exemplary embodiment of the present invention's design.
Service load receiver clock output signal (O_RCLK) G1 produced by receiver clock generator 62 with reference to Fig. 4, G1.By multiple impact damper Bu50, Bu51 and the Bu52 on the 5th bus line B22, service load receiver clock output signal (O_RCLK) G1 is sent to first interface 120 from the second interface 220.
G2 is the increment signal by exporting the latch that control signal generator 52 produces based on indicator signal G2A, to increase local read pointer G3.
G3 is local read pointer (RPTR_LCL), increases when local read pointer in a period of time (RPTR_LCL) of the entrance in the FIFO memory 34 that service load receiver 240 reads in the second interface 220.
G4 is that the response of reading produced by trigger 60 outputs signal (O_RACK).
G5 is that reading in first interface 120 replys input signal (I_RACK), and it is the delay version that reading in the second interface 220 replys output signal (O_RACK) G4.Reading in the second interface 220 is replied output signal (O_RACK) G4 and is propagated by impact damper Bu40, Bu41 and the Bu42 on the 4th bus line B21 in long hop channel B10, and is input to the trigger 20 in first interface 120.
G6 is the receiver clock input signal (I_RCLK) in first interface 120, and it is the delay version of service load receiver clock output signal (O_RCLK) G1 in the second interface 220.Service load receiver clock output signal (O_RCLK) G1 in second interface 220 is propagated by impact damper Bu50, Bu51 and the Bu52 on the 5th bus line B22 in long hop channel B10, and is input to the clock recovery unit 26 in first interface 120.
G7 be based on read reply input signal (I_RACK) G5 with recover receiver clock RCLK1 synchronous read answer signal.
G8 is the long-range read pointer (RPTR_RMT) produced by the enable signal generator 22 in first interface 120.When reading increment signal G7B and activating, long-range read pointer G8 increases.
The 0-6 of G3 corresponds to local read pointer value, and the 0-6 of G8 corresponds to long-range read pointer value.
Service load receiver clock can be outputed signal (O_RCLK) G1 and be provided to trigger 54 and 60 in the service load receiver 240 of the second interface 220 in Fig. 2.Receiver clock input signal (I_RCLK) G6 can be provided to the trigger 20 in the receiver 140 of the first interface 120 in Fig. 2 and long-range read pointer generator 24.If receiver clock input signal G6 had controlled clock skew before being provided to trigger 20 and long-range read pointer generator 24, then recover receiver clock input signal G6 by clock recovery unit 26.
Signal G1 and G4 can start from the second interface 220, and is delayed by such as more than a clock period, shown in the arrow A R2 in Fig. 4.Delay shown in arrow A R2 can be due to the layout routing delay based on temperature or route conditions, and described route conditions is such as the length of the route conductors of the layout route in SoC 150, width, resistance or stray capacitance.
When the second interface 220 is bide one's time before one of entrance that can read FIFO memory 34 etc. at service load receiver 240, service load receiver clock output signal (O_RCLK) G1 can stop triggering; In addition, service load receiver clock output signal (O_RCLK) G1 can always trigger.First interface 120 by recovering service load receiver clock output signal (O_RCLK) G1 with under type, that is, produces the clock RCLK1 recovered from the clock input signal received (I_RCLK) G6 through the 5th bus line B22.
Long hop channel L20 can perform synchronized transaction, is synchronously sent to service load reservoir 230 the second interface 220 by writing the transmitter 130 of data output signal (O_WDATA) F6 from first interface 120 to output signal (O_TCLK) F1 with transmitter clock.Short channel L10 can operation exception affairs, service load to be sent to the multiplexer 37 in the second interface 220 from the service load reservoir 230 the second interface 220.
Long hop channel L20 can perform synchronized transaction, synchronously outputs signal reading to reply (O_RACK) G4 and to be sent to receiver 140 first interface 120 from the service load receiver 240 of the second interface 220 to output signal (O_RCLK) G1 with service load receiver clock.Short channel L11 can operation exception affairs, long-range read pointer (RPTR_RMT) G8 to be sent to the synchronizer 16 in first interface 120 from the receiver 140 first interface 120.
According to the description of Fig. 1 to Fig. 4, compared with having the asynchronous interface of long asynchronous transmission line, the change from long asynchronous transmission line to long synchronous transmission line can make the interface of the long hop channel had transmitter circuit 100 and acceptor circuit 200 simple and efficient.
With reference to Fig. 1, as the 2nd IP 260, as main IP operation, the one IP 160 is as when operating from IP, and long hop channel B20 can operate identically with long hop channel B10.In this case, the second interface 220 can comprise transmitter and receiver.First interface 120 can comprise service load reservoir and service load receiver.
If SoC 150 comprises long hop channel B10 and B20, then an IP 160 and the 2nd IP 260 can perform bidirectional operation to transmit service load.
Fig. 5 is the block diagram of asynchronous interface circuit of the exemplary embodiment according to the present invention's design.
With reference to Fig. 5, asynchronous interface circuit 115 can comprise first interface 124, second interface 222 and two-way long hop channel.Two-way long hop channel can comprise the first long hop channel B10 and the second long hop channel B20.
First interface 124 can comprise the first transmitter 130, reads unit 140 as first of receiver, as the second r/w cell 232 and the second receiver 242 of service load reservoir.
Second interface 222 can comprise the second transmitter 132, reads unit 142 as second of receiver, as the first r/w cell 230 and the first receiver 240 of service load reservoir.
First long hop channel B10 can comprise multiple bus circuit B11, B12, B13, B21 and B22.Described multiple bus circuit B11, B12 and B21 in Fig. 5 may correspond to bus line B11, B12 and B21 in Fig. 2.Conveniently explain, the bus line B13 in Fig. 2 and bus line B22 can send first interface clock signal clk 1 in Fig. 5 and the second interface clock signal CLK2 respectively.
Second long hop channel B20 can comprise multiple bus line B11-1, B12-1, B13, B21-1 and B22.Multiple bus circuit B11-1, B12-1 in Fig. 5 and B21-1 can be identical with B21 with bus line B11, the B12 in Fig. 5 operate.
Therefore, the first long hop channel B10 and the second long hop channel B20 can shared bus circuit B13 and B22, to transmit first interface clock signal clk 1 and the second interface clock signal CLK2.In other words, long hop channel B10 and B20 of shared bus circuit B13 and B22 can perform bidirectional operation, to transmit service load independently.
And in the second interface 222 second reads unit 142 and the first r/w cell 230 can share first interface clock signal clk 1 as clock source.The second r/w cell 232 and first in first interface 124 reads unit 140 can share the second interface clock signal CLK2 as clock source.
First service load can be sent to the first r/w cell 230 by the first transmitter 130 in first interface 124, and first in first interface 124 is read unit 140 and can be received answer signal from the first receiver 240 the second interface 222.
Second service load can be sent to the second r/w cell 232 by the second transmitter 132 in the second interface 222, and second in the second interface 222 is read unit 142 and can be received answer signal from the second receiver 242 first interface 124.
According to Fig. 5, at least one in first interface 124 and the second interface 222 can be used as main interface or from interface operation.
Fig. 6 is the block diagram of SoC of the exemplary embodiment according to the present invention's design.
Can comprise as by the transmitter circuit 100a of at least one functional block be connected in long hop channel B1 with B2 and acceptor circuit 200a with reference to Fig. 6, SoC 151.Although named transmitter circuit 100a and acceptor circuit 200a for convenience of description, transmitter circuit 100a and acceptor circuit 200a has named interchangeably.
Transmitter circuit 100a can comprise as the IP 160a from IP and the first interface 124a as asynchronous main interface AMI.
Acceptor circuit 200a can comprise as the 2nd IP 260a of main IP with as asynchronous the second interface 222a from interface ASI.
Each of long hop channel B1 and B2 can comprise effective load signal, write enable signal, transmitter clock signal, receiver clock signal and read answer signal.Channel B1 and B2 can share transmitter clock signal and receiver clock signal.
First interface 124a can comprise buffer unit 232a, synchronously to receive service load from the second interface 222a with the transmitter clock produced by the second interface 222a.Buffer unit 232a can comprise FIFO memory.
Second interface 222a can comprise buffer unit 230a, synchronously to receive service load from first interface 124a with the transmitter clock produced by first interface 124a.Buffer unit 230a can comprise FIFO memory.
One IP 160a is connected to first interface 124a by bus B 1a and B2b.One IP 160A can comprise the channel 162 (CHSA) and 164 (CHSB) that can be respectively and read channel or write-channel.Channel 162 (CHSA) communicates with first interface 124a with B2a by bus B 1a with 164 (CHSB).
One IP 160a can be the memory assembly with the multiple channels be connected with B2b by bus B 1a.Each of multiple channel 162 (CHSA) and 164 (CHSB) can be the one-way channel for reading or writing, or can be the two-way channel for read and write.Here, memory assembly can be register, volatile memory (such as SRAM or DRAM), nonvolatile memory (such as NAND flash memory, NOR flash memory, phase change random access memory devices (PRAM), ferroelectric RAM (FRAM)) etc.
2nd IP 260a is connected to the second interface 222a by bus B 1b and B2a.2nd IP 260a can comprise the channel 261 (CHMA) and 262 (CHMB) that can control respectively from IP.Channel 261 (CHMA) communicates with the second interface 222a with B2a by bus B 1b with 262 (CHMB).
2nd IP 260a can be the Memory Controller with the multiple channels be connected with B2a by bus B 1b.Each of multiple channel 261 (CHMA) and 262 (CHMB) can be the one-way channel read or write for control store assembly.
Fig. 7 is the process flow diagram of the method illustrated according to the first asynchronous interface in the application drawing 2 of the exemplary embodiment of the present invention's design.
With reference to Fig. 2, the method operating the first asynchronous interface can comprise step: sent by first interface 120 and write data output signal (O_WDATA) F6, write enable output signal (O_WEN) F5 and transmitter clock output signal (O_TCLK) F1.
Referring now to Fig. 2 and Fig. 7, in step S710, the second interface 220 is configured to arrange service load reservoir 230 to receive service load from first interface 120.
In step S720, the second interface 220 receives service load F6 by bus line B11 from first interface 120, and latches service load F6 by the trigger 30 in service load reservoir 230.In addition, the second interface 220 receives by bus line B12 and B13 and writes enable output signal F5 and transmitter clock output signal F1.Bus line B11, B12 and B13 can in long hop channel L20, and described long hop channel L20 has long conductor circuit between first interface 120 and the second interface 220.
In step S730, clock recovery unit 38 by adjusting clock skew between the service load that receives and the transmitter clock signal received to produce the transmitter clock signal TCLK1 of recovery, thus makes clock skew allow within the scope of clock skew maximum.Clock recovery unit 38 can comprise the DLL circuit for adjusting clock skew.
In step S740, based on the remote write pointer F12 produced in the second interface 220 service load received is stored in the FIFO memory 34 in service load reservoir 230.Remote write pointer F12 can be destination address, to select in each entrance of FIFO memory 34 based on the transmitter clock signal TCLK1 recovered and the write enable signal received.
In step S750, by short channel L10, the service load be stored in FIFO memory 34 is sent to service load receiver 240 asynchronously.
Therefore, the method operating the first asynchronous interface can make the first asynchronous interface between functional block far away apart tell in circuit complexity and conveying function.
Fig. 8 is the process flow diagram of the method illustrated according to the second asynchronous interface in the application drawing 2 of the exemplary embodiment of the present invention's design;
With reference to Fig. 2, the method operating the second asynchronous interface can comprise step: receive receiver clock input signal (I_RCLK) G6 by first interface 120 and read to reply input signal (I_RACK) G5.
Referring now to Fig. 2 and Fig. 8, in step S810, first interface 120 receives receiver clock input signal G6 by long hop channel L20 from the second interface 220 and replys input signal G5 with reading.
In step S820, clock recovery unit 26 by adjusting clock skew between the answer signal that receives and the receiver clock signal received to produce the receiver clock signal RCLK1 of recovery, thus makes clock skew allow within the scope of clock skew maximum.Clock recovery unit 26 can comprise the DLL circuit for adjusting clock skew.
In step S830, the trigger 20 in first interface 120 can based on read reply input signal (I_RACK) G5 produce with recovery receiver clock signal RCLK1 synchronous read answer signal G7.Enable signal generator 22 can produce read increment signal G7B based on the answer signal G7 that reads synchronous with the receiver clock signal RCLK1 recovered.
In step S840, be sent to transmitter 130 in first interface 120 by short channel L11 asynchronously by reading increment signal G7B.
Therefore, the method operating the second asynchronous interface can make the second asynchronous interface between functional block far away apart tell in circuit complexity and conveying function.
When achieving the exemplary embodiment of the present invention's design of Fig. 1 to Fig. 6 in SoC, static timing analysis (STA) can be used to stop (s i gn-off) condition to meet Digital Logic sequential.Designer's adjustable distribution element floor plan (layout floor planning) long hop channel L20 and short channel L10 and L11 to be arranged in the tram in the layout of SoC, to meet the design specification for speed and power consumption.
When the clock skew between transmitter clock signal F9 and the service load F8 received received can be adjusted in maximum allowable range, the transmitter clock signal F9 received can be used as the tranmitting data register signal recovered, and need not perform clock recovery process.
When the clock skew between receiver clock signal G6 and the answer signal G5 received received can be adjusted in maximum allowable range, the receiver clock signal G6 received can be used as the receiver clock signal recovered, and need not perform clock recovery process.
Fig. 9 is the block diagram of SoC of the exemplary embodiment according to the present invention's design.
Multiple main IP 600,601 and 602, network-on-chip (on-chip network) 500, multiple from IP 700 and 701 and memory interleaving device (hereinafter referred to as MID) 105 and 106 can be comprised with reference to Fig. 9, SoC 1000.
Main IP 600,601 and 602 can be CPU, encoder (CODEC), display, imageing sensor etc.Memory mapped device is can be from IP 700 and 701.
Network-on-chip 500 can be the network interface for management data and control flow check in SoC 1000.Network-on-chip 500 can be realized in the same substrate of SoC 1000 or realize on more than one chip.
Each and two that each of MID 105 and 106 can be connected to three main IP 600,601 and 602 from IP 700 and 701 each between, and based on control information, read/write requests can be dispensed to from IP 700 and 701 from main IP 600 to 602.Three main IP 600,601 and 602 can be connected to each of MID 105 and 106 by network-on-chip 500.
Can according to Advanced extensible Interface (MAXI) the bus protocol operation of amendment with reference to Fig. 9, SoC 1000.In other words, can use as transmitter clock and receiver clock provide the asynchronous handshake feature of service load and response signal, effectively operate in long hop channel to make bus protocol.Here, in order to the effect of performance and complexity aspect, the FIFO memory of sender side is moved to receiver-side.And, this asynchronous interface can be applicable to other bus protocol, the advanced system bus (ASB) etc. in such as AXI, Advanced High-Performance Bus (AHB), advanced peripheral bus (APB) and Advanced Microcontroller Bus Architecture (AMBA).By increasing any one in each signal of such as transmitter clock, receiver clock, enable signal and answer signal, can be the bus protocol of the asynchronous bus interface contained in long hop channel by any synchronous bus protocol modification.
When in main IP 600 to 602 needs access from IP 700 and 701 one, wherein accessedly can operate in the clock zone different from the clock zone of in main IP 600 to 602 from IP 700 and 701, each of MID 105 and 106 can according to MAXI protocol operation.
Two functional blocks operated under AXI agreement in different clock zones (such as, first main IP 600 and first is from IP 700) between asynchronous interface, another by described two functional blocks clock signal being separately sent in described two functional blocks, described two functional blocks can operate under MAXI agreement.
The quantity being connected to the MAXI channel of network-on-chip 500 or MID 105 and 106 is not limited to two or three, but can according to being connected to the main IP of network-on-chip 500 or MID 105 and 106 and being one or more than three from the quantity of IP.
When in MAXI channel is long hop channel, the asynchronous interface circuit 105 in Fig. 2 can be the MID 105 and/or 106 of the SoC 1000 in Fig. 9.Asynchronous interface circuit 105 in Fig. 2 can be included in the network-on-chip 500 of SoC 1000 in fig .9.
Figure 10 is the block diagram of memory interleaving device of the exemplary embodiment according to the present invention's design.
Three can be comprised from interface 220 (SI0), 221 (SI 1), 222 (SI2), two main interfaces 120 (MI0) and 121 (MI 1) and cross bar switches 140 with reference to Figure 10, MID 105.Interface 120 (MI0), 121 (MI 1), 220 (SI0), 221 (SI 1) and 222 (SI2) can according to MAXI protocol operation.
For convenience of description, based on the entitlement of data stream, the interface being connected to main IP is called from interface, the interface be connected to from IP is called main interface.
Network-on-chip 500 can be connected to from interface 220 (SI0), 221 (SI 1) and 222 (SI2).Main interface 120 (MI0) and 121 (MI 1) can be connected to from IP 700 and 701.Be connected to each other from interface 220 (SI0), 221 (SI 1) and 222 (SI2) and main interface 120 (MI0) and 121 (MI 1) by cross bar switch 140.
Connect from (such as, 120 (MI0)) in interface (such as, 220 (SI0)) and main interface by asynchronous long hop channel.The FIFO memory for storing from the read data sent from IP 700 can be comprised from interface 220 (SI0).Main interface 120 (MI0) can comprise for storing the FIFO memory writing data sent from (Fig. 9's) main IP 600.
Asynchronous long hop channel is connected between interface 220 (SI0) and main interface 120 (MI0) by cross bar switch 140.The asynchronous long hop channel of MAXI bus protocol design described by Fig. 1 to Fig. 9.
Although by performing from the read/write data transmission between interface 220 (SI0) and main interface 120 (MI0) with providing the clock signal synchronization synchronous with read/write data, perform (Fig. 9's) main IP 600 and (Fig. 9's) asynchronously from the read/write data transmission between IP 700.As described in Fig. 1 to Fig. 9, the read/write data received is sent to main IP 600 by short channel or asynchronously from IP 700.
Clock signal is provided by providing the transmitter reading or writing data and answer signal.Answer signal can send between interface 220 (SI0) and main interface 120 (MI0) in response to read/write data, and asynchronous with read/write data.
Cross bar switch 140 can distribute the read/write requests and main IP 600 to 603 initiated by main IP 600 to 603 and from the read/write data between IP 700 to 701.
Therefore, by utilizing AXI bus protocol in the SoC 1000 that describes in figure 9 and in figure 10 and having the network-on-chip of MID 105 and 106, the asynchronous interface circuit 105 in Fig. 2 can be applicable to SoC 1000.
Figure 11 is the block diagram of multiple MAXI channels of the exemplary embodiment according to the present invention's design.
MAXI bus protocol comprises reading address channel (hereinafter referred to as AR-channel) and read data channel (hereinafter referred to as R-channel), write address channel (hereinafter referred to as AW-channel), write data channel (hereinafter referred to as W-channel) and write response channel (hereinafter referred to as B-channel) for read operation.
AR-channel can be sent to having the address (ARADDR) of reading of reading address enable signal (ARVALID) from machine from main frame.Then, answer signal (ARREADY) can be sent to main frame from from machine by AR-channel.
The read data (RDATA) with read data enable signal (RVALID) can be sent to main frame from from machine by R-channel.Then, answer signal (RREADY) can be sent to from machine from main frame by R-channel.
The write address (AWADDR) with write address enable signal (AWVALID) can be sent to from machine from main frame by AW-channel.Then, answer signal (AWREADY) can be sent to main frame from from machine by AW-channel.
The data (WDATA) of writing with write enable signal (WVALID) can be sent to from machine from main frame by W-channel.Then, answer signal (WREADY) can be sent to main frame from from machine by W-channel.
The response (BRESP) write with response enable signal (BVALID) can be sent to main frame from from machine by B-channel.Then, answer signal (BREADY) can be sent to from machine from main frame by B-channel.
AW-channel can comprise AWID (write address ID), AWADDR (write address), AWLEN (pulse length), AWSIZE (impulse magnitude), AWBURST (pulse pattern), AWVALID (write address/control is effective) and AWREADY (write address/control accepts).
W-channel can comprise WID (writing data ID), WDATA (writing data), WSTRB (write gate), WLAST (last in pulse writes transmission), WVALID (writing data effective) and WREADY (writing data to accept).
B-channel can comprise BID (writing data ID), BRESP (write response), BVALID (write response is effective) and BREADY (write response acceptance).
AR-channel can comprise ARID (reading address ID), ARADDR (reading address), ARLEN (pulse length), ARSIZE (impulse magnitude), ARBURST (pulse pattern), ARVALID (reading address/control effective) and ARREADY (reading address/control to accept).
R-channel can comprise RID (read data ID), RDATA (read data), RRESP (reading to respond), RLAST (last in pulse is read to transmit), RVALID (read data is effective) and RREADY (read data acceptance).
Each of AR-channel, R-channel, AW-channel, W-channel and B-channel can be asynchronous long hop channel, and can be separately located in main frame and between machine, and can share host clock MI_CLK and from machine clock SI_CLK.
Therefore, AR-channel can be connected to FIFO memory (such as, reading address buffer queue), with store sent by AR-channel read address.R-channel can be connected to FIFO memory (such as, read data buffer queue), to store the read data sent by R-channel.AW-channel can be connected to FIFO memory (such as, write address buffer queue), to store the write address sent by AW-channel.W-channel can be connected to FIFO memory (such as, writing data buffer queue), with store sent by W-channel write data.B-channel can be connected to FIFO memory (such as, response buffering sector arranges), to store the response sent by B-channel.The details that Figure 12 will provide about the FIFO memory as buffer queue.
Useful signal in the channel of MAXI agreement and ready signal may correspond to enable signal in the channel in Fig. 1 and Fig. 6 and answer signal.Such as, the bus line B11 in Fig. 2 and bus line B21 can correspond respectively to WDATA and WREADY of W-channel.
Asynchronous long hop channel can comprise for the AW-channel of write operation, W-channel and B-channel, and for the AR-channel of read operation and R-channel.
Asynchronous long hop channel can comprise and having for read operation and the host clock MI_CLK of write operation and the clock channel CK-channel from machine clock SI_CLK.Such as, the bus line B11 in Fig. 2 and bus line B21 can correspond respectively to W-channel and B-channel.
Figure 12 illustrates according to the main interface in the SoC of the exemplary embodiment of the present invention's design and the block diagram from the affairs between interface.
SoC can comprise by AR-channel, R-channel, AW-channel, W-channel or B-channel connect from interface 220 and main interface 120.
Main interface 120 can comprise reads address FIFO memory AR-FIFO for AR-channel, to read address from receiving from interface 220.The read data FIFO memory R-FIFO for R-channel can be comprised, to receive read data from main interface 120 from interface 220.
Main interface 120 also can comprise the write address FIFO memory AW-FIFO for AW-channel, to receive write address from from interface 220.Main interface 120 also can comprise writes data fifo memory W-FIFO for W-channel, to write data from receiving from interface 220.The response FIFO memory B-FIFO for B-channel also can be comprised, to receive response from main interface 120 from interface 220.
With reference to Figure 12, at main interface 120 and between interface 220, read affairs by utilizing AR-channel (CH1) and R-channel (CH2) to perform, and write affairs by utilizing AW-channel (CH10), W-channel (CH20) and B-channel (CH30) to perform.
Read affairs and write affairs can comprise there is host clock MI_CLK and the CK-channel from machine clock SI_CLK.AR-channel, R-channel, AW-channel, W-channel and B-channel can share host clock MI_CLK and from machine clock SI_CLK.
Reading in affairs, address will read from interface 220 by AR-channel and control information is sent to main interface 120, to send read request.Read address to store or queuing in the AR-FIFO in main interface 120.In response to read request, read data is sent to from interface 220 by R-channel by main interface 120.Read data can store or queue up from the R-FIFO in interface 220.
Reading in affairs, each of AR-channel and R-channel can be asynchronous long hop channel.
With reference to Figure 11 and Figure 12, the ARVALID signal as enable signal and the ARADDR signal as service load are synchronously sent to main interface 120 from from interface 220 by AR-channel and host clock MI_CLK.As answer signal ARREADY signal by AR-channel be synchronously sent to from interface 220 from main interface 120 from machine clock SI_CLK.AR-channel can utilize ARADDR signal to send control information.ARADDR signal, ARVALID signal and ARREADY signal can correspond respectively to bus line B11, B12 and B21 in Fig. 2.Host clock MI_CLK and bus line B22 and B13 that can correspond respectively to from machine clock SI_CLK Fig. 2.
RVALID signal as enable signal and the RDATA signal as service load are by R-channel and be synchronously sent to from interface 220 from main interface 120 from machine clock SI_CLK.RREADY signal as answer signal is synchronously sent to main interface 120 from from interface 220 by R-channel and host clock MI_CLK.RDATA signal, RVALID signal and RREADY signal can correspond respectively to bus line B11, B12 and B21 in Fig. 2.Host clock MI_CLK and bus line B22 and B13 that can correspond respectively to from machine clock SI_CLK Fig. 2.
On the other hand, asynchronous long hop channel can comprise for the AR-channel of read operation, R-channel and CK-channel.
AR-channel can synchronously send the service load (such as, ARADDR, ARID, ARSIZE etc.) with enable signal (such as, ARVALID or AREN) with the host clock MI_CLK of CK-channel.R-channel can with synchronously send the answer signal (such as, RVALID or RSTRB) with read data (such as, RDATA) from machine clock SI_CLK.
Writing in affairs, each of AW-channel, W-channel and B-channel can be asynchronous long hop channel.
With reference to Figure 11 and Figure 12, the AWVALID signal as enable signal and the AWADDR signal as service load synchronously can be sent to main interface 120 from from interface 220 by AW-channel and host clock MI_CLK.Can using as answer signal AWREADY signal by AW-channel be synchronously sent to from interface 220 from main interface 120 from machine clock SI_CLK.AW-channel can utilize AWADDR signal to send control information.AWADDR signal, AWVALID signal and AWREADY signal can correspond respectively to bus line B11, B12 and B21 in Fig. 2.Host clock MI_CLK and bus line B22 and B13 that can correspond respectively to from machine clock SI_CLK Fig. 2.
WVALID signal as enable signal and the WDATA signal as service load synchronously can be sent to main interface 120 from from interface 220 by W-channel and host clock MI_CLK.Can using as answer signal WREADY signal by W-channel be synchronously sent to from interface 220 from main interface 120 from machine clock SI_CLK.W-channel can utilize WADDR signal to send control information.WADDR signal, WVALID signal and WREADY signal can correspond respectively to bus line B11, B12 and B21 in Fig. 2.Host clock MI_CLK and bus line B22 and B13 that can correspond respectively to from machine clock SI_CLK Fig. 2.
Can using the BVALID signal as enable signal and the BRESP signal as service load by B-channel be synchronously sent to from interface 220 from main interface 120 from machine clock SI_CLK.BREADY signal as answer signal synchronously can be sent to main interface 120 from from interface 220 by B-channel and host clock MI_CLK.BRESP signal, BVALID signal and BREADY signal can correspond respectively to bus line B11, B12 and B21 in Fig. 2.Host clock MI_CLK and bus line B22 and B13 that can correspond respectively to from machine clock SI_CLK Fig. 2.
On the other hand, asynchronous long hop channel can comprise for the AW-channel of read operation, W-channel, B-channel and CK-channel.
AW-channel can synchronously send first service load (such as, AWADDR, AWID, AWSIZE etc.) with the first enable signal (such as, AWVALID, AWEN) with the host clock MI_CLK of CK-channel.W-channel can synchronously send second service load (such as, WDATA, WID, WSTRB etc.) with the second enable signal (such as, WVALID or WEN) with the host clock MI_CLK of CK-channel.B-channel can synchronously send answer signal (such as, BRESP) with from machine clock SI_CLK.
Figure 13 is the block diagram of SoC of the exemplary embodiment according to the present invention's design.
Multiple main frame 260a, multiple from machine 160a be connected to described multiple main frame 260a and described multiple bus system 105a from machine 160a can be comprised with reference to Figure 13, SoC 1300.SoC 1300 can be embodied as the chip comprised in an enclosure.
Multiple main frame 260a can comprise the first main frame 20-1, the second main frame 20-2, the 3rd main frame 20-3 and the 4th main frame 20-4.Multiplely can comprise first from machine 40-1, second from machine 40-2, the 3rd from machine 40-3 and the 4th from machine 40-4 from machine 160a.In order to effectively explain, main frame and the quantity from machine are four, but main frame and the quantity from machine are not limited thereto.
Bus system 105a can comprise priority controller 31 and bus switch 33.Priority controller 31 can control the priority that multiple main frame 260a is connected with multiple interface between machine 160a.Priority controller 31 from multiple main frame 260a and multiple multiple bus request being used for read/write operation from machine 160a reception, and can manage the priority of serving each bus request.
Bus switch 33 can comprise first from interface 33-1 (SI1), second from interface 33-2 (SI2), the 3rd from interface 33-3 (SI3), the 4th from interface 33-4 (SI4), the first main interface 33-5 (MI1), the second main interface 33-6 (MI2), the 3rd main interface 33-7 (MI3) and the 4th main interface 33-8 (MI4).
At least one in multiple main frame 260a and can be connected to multiple from least one machine 160a by the destination address produced of in main frame 260a based on the precedence information produced by priority controller 31 by bus switch 33.Bus switch 33 by be connected to selection in multiple main frame 260a at least one from one of interface 33-1 to 33-4 and be connected to one of multiple main interface 33-5 to 33-8 of at least one from the selection machine 160a at least one of the selection in multiple main frame 260a be connected to multiple at least one from the selection machine 160a.At multiple main frame and multiple from uncontested between machine or when conflicting, it is multiple from machine that bus switch 33 can support multiple main frame to access simultaneously.
Can according to AMBA3 or AMBA4 agreement or bus protocol design bus subsystem of shaking hands.
According to the exemplary embodiment of the present invention's design in Figure 13, the main frame 260a in bus system 105a and can be asynchronous long hop channel from least one in the connection (dotted line Figure 13) between machine 160a.By providing the main frame MI_CLK that describes in Fig. 9 and Figure 12 and from machine clock SI_CLK, with regard to circuit complexity, performance and power consumption, can design bus switch 33 effectively, as described in Fig. 1 and Figure 12.
Each of main frame 20-1 to 20-4 can be the microprocessor or graphic process unit that realize in SoC 1300.SoC 1300 can be integrated circuit, and can realize in the various mobile devices of such as mobile phone, smart phone, tablet personal computer (PC), personal digital assistant (PDA) etc.SoC 1300 can realize in infotech (IT) device or portable electron device.
Figure 14 is the block diagram of the data handling system of the SoC of the exemplary embodiment comprised according to the present invention's design.
With reference to Figure 14, data handling system 2000 can comprise SoC 150, antenna 201, radio frequency (RF) transceiver 203, input media 205 and display 207.SoC 150 can be the SoC 150 shown in Fig. 1.
RF transceiver 203 receives by antenna 201 and sends wireless signal.The wireless signal received can be converted to the signal that can be processed by SoC 150 by RF transceiver 203.
SoC 150 can process the signal exported from RF transceiver 203, and treated signal is sent to display 207.In addition, the signal produced by SoC 150 can be converted to wireless signal by RF transceiver 203, and wireless signal is sent to external device (ED) by antenna 201.
Data or the input media 205 be used in the control information input SoC 150 of control SoC 150 be can be the point apparatus of such as touch pad, computer mouse, keypad or keyboard.
Data handling system 2000 can comprise the asynchronous long hop channel in SoC 150, and can reduce design complexities and power consumption.
Figure 15 is the block diagram of the data handling system of the SoC of the exemplary embodiment comprised according to the present invention's design.
With reference to Figure 15, data handling system 3000 can be realized in PC, the webserver, dull and stereotyped PC, net book, electronic reader, PDA, portable media player (PMP), MP3 player or MP4 player.
Data handling system 3000 can comprise SoC 150, storage arrangement 301, Memory Controller 302, display 303 and input media 304 for the data processing of control store apparatus 301.
Input signal can be converted to data by input media 304, and data is sent to SoC150 or Memory Controller 302.
SoC 150 can receive the data inputted from input media 304.Data can be shown or be stored in storage arrangement 301 under the control of SoC 150.The data be stored in storage arrangement 301 can be shown by display 303 under the control of Memory Controller 302.
SoC 150 can the operation of overall control data disposal system 3000 diode-capacitor storage controller 302.Memory Controller 302 can realize in SoC 150, or is designed to the assembly of separation.
According to the exemplary embodiment of the present invention's design in Fig. 1 to Figure 13, data handling system 3000 can comprise the asynchronous long hop channel in SoC 150, and can reduce design complexities and power consumption.
Figure 16 is the block diagram of the data handling system of the SoC of the exemplary embodiment comprised according to the present invention's design.
With reference to Figure 16, data handling system 4000 can realize at such as digital camera, have in the mobile phone of camera model or the image processing apparatus of smart phone.
Data handling system 4000 can comprise SoC 150, storage arrangement 401, Memory Controller 402, display 404 and imageing sensor 403 for the data processing of control store apparatus 401.
Optical image data can be converted to Digital Image Data by imageing sensor 403, and Digital Image Data is sent to SoC 150 or Memory Controller 402.
SoC 150 can receive the Digital Image Data inputted from imageing sensor 403.Digital Image Data can be shown or be stored in storage arrangement 401 under the control of SoC 150.The Digital Image Data be stored in storage arrangement 401 can be shown by display 404 under the control of Memory Controller 402.
SoC 150 can the operation of overall control data disposal system 4000 diode-capacitor storage controller 402.Memory Controller 402 can realize in SoC 150, or is designed to the assembly of separation.
According to the exemplary embodiment of the present invention's design in Fig. 1 to Figure 13, data handling system 4000 can comprise the asynchronous long hop channel in SoC 150, and can reduce design complexities and power consumption.
Figure 17 is the block diagram of the computer system of the SoC of the exemplary embodiment comprised according to the present invention's design.
With reference to Figure 17, computer system 5000 can comprise multiple non-volatile memory device 501, storage arrangement 503, for the Memory Controller 502 of the operation of control store apparatus 501 and storage arrangement 503 with for the data processed by storage arrangement 501 and main frame 504 being stored in the SoC 150 in storage arrangement 503.
Multiple non-volatile memory devices 501 can be the nonvolatile memory of such as nand memory and NOR storer.Storage arrangement 503 can comprise the volatile memory of such as DRAM and SRAM or the nonvolatile memory of such as magnetic random access memory (MRAM).
Memory Controller 502 can be connected with external device interface according to communication protocol, and described communication protocol is USB (universal serial bus) (USB), multimedia card (MMC), periphery component interconnection (PCI), high-speed PCI (PCI-E), Advanced Technology Attachment (ATA), serial ATA, Parallel ATA, small computer system interface (SCSI), enhancing minidisk interface (ESDI) and integrated driving electronic equipment (IDE) such as.
Multiple non-volatile memory device 501 is connected to Memory Controller 502 by multiple memory channel.Each of described multiple non-volatile storage component 501 can realize according to lower device: NAND flash memory, Electrically Erasable Read Only Memory (EEPROM), MRAM, spin-transfer torque MRAM, conductive bridge RAM (CBRAM), ferroelectric RAM (FeRAM), be referred to as the PRAM of Ovshinsky Unified Memory (OUM), resistance-type RAM (RRAM or ReRAM), nanotube RRAM, polymkeric substance RAM (PoRAM), nanometer floating gate memory (NFGM), holographic memory, molecular electronic storage arrangement or insulator Resistance-change memory.
Each of multiple non-volatile memory device 501 and storage arrangement 503 can be encapsulated in various packaging part, described packaging part is laminate packaging (PoP) such as, ball grid array (BGA), chip size packages (CSP), plastic leaded chip carrier (PLCC), plastics dual-in-line package (PDIP), China husband device wafer, China husband form chip, chip on board (COB), ceramic dual in-line package (CERDIP), plastics metric system quad flat package (MQFP), plastic quad flat package (PQFP), small outline integrated circuit (SOIC), the little outline packages of shrinkage type (SSOP), Thin Small Outline Package (TSOP), slim quad flat package (TQFP), system in package (SIP), multi-chip package (MCP), wafer scale manufacturing and encapsulation (WFP) or wafer-level process laminate packaging (WSP) etc.
According to the exemplary embodiment of the present invention's design in Fig. 1 to Figure 13, computer system 5000 can comprise the asynchronous long hop channel in SoC 150, and can reduce design complexities and power consumption.
Computer system 5000 can realize with in lower device: super mobile PC (UMPC), workstation, net book, PDA, portable computer, web-tablet, flat computer, wireless phone, mobile phone, smart phone, e-book, PMP, portable game machine, navigational system, black box, digital camera, DMB (DMB) player, three-dimensional television, digital audio frequency recording instrument, digital audio-frequency player, digital picture registering instrument, digital picture player, digital video recorder, video frequency player, for the storage medium of data center, wireless transceiver/receiver system, for home network, computer network, one of the various electronic installation of teleprocessing network or assembly, radio-frequency (RF) identification (RFID) device, computing system etc.
Although the exemplary embodiment with reference to the present invention's design describes the present invention's design, but it should be understood by one skilled in the art that, when not departing from the spirit and scope of the present invention's design be defined by the claims, the amendment in various forms and details can be made to it.

Claims (20)

1. a System on Chip/SoC, comprising:
First interface;
Second interface;
3rd interface;
First channel, it is connected between described first interface and described second interface;
Second channel, it is connected between described first interface and described 3rd interface; And
Clock channel, it has the first clock signal, second clock signal and the 3rd clock signal, described first clock signal is connected between at least one in described first interface and described second interface and the 3rd interface, described second clock signal is connected between described first interface and described second interface, and described 3rd clock signal is connected between described first interface and described 3rd interface.
2. System on Chip/SoC according to claim 1, also comprises:
First main intellecture property, it is connected to described first interface;
First from intellecture property, and it is connected to described second interface; And
Second from intellecture property, and it is connected to described 3rd interface.
3. System on Chip/SoC according to claim 1, wherein, described first channel and described second channel are according to the communication protocol operations based on advanced extensible interface bus agreement.
4. System on Chip/SoC according to claim 3, wherein, by the first data-signal and the first enable signal and described first clock signal synchronization from described first interface be sent to described second interface and described 3rd interface at least one.
5. System on Chip/SoC according to claim 4, wherein, second data-signal and the second enable signal and described second clock signal are synchronously sent to described first interface from described second interface, or the 3rd data-signal and the 3rd enable signal and described 3rd clock signal synchronization are sent to described first interface from described 3rd interface.
6. System on Chip/SoC according to claim 5, wherein, by the first answer signal and described first clock signal synchronization from described first interface be sent to described second interface and described 3rd interface at least one.
7. System on Chip/SoC according to claim 6, wherein, second answer signal and described second clock signal are synchronously sent to described first interface from described second interface, or the 3rd answer signal and described 3rd clock signal synchronization are sent to described first interface from described 3rd interface.
8. System on Chip/SoC according to claim 1, wherein, described first interface comprises local write pointer generator, and at least one in described second interface and described 3rd interface comprises remote write pointer generator.
9. System on Chip/SoC according to claim 8, wherein, at least one in described second interface and described 3rd interface comprises local read pointer generator, and described first interface comprises long-range read pointer generator.
10. a bus interface method of attachment for System on Chip/SoC, comprises step:
With the first transfer rate, service load and the first clock signal synchronization are sent to buffer memory the second interface by the first channel from first interface; And
With the second transfer rate, described service load is sent to service load receiver from described buffer memory by the second channel asynchronous with described first channel, the length of wherein said first channel is longer than the length of described second channel.
The bus interface method of attachment of 11. System on Chip/SoCs according to claim 10, wherein, described buffer memory is push-up storage.
The bus interface method of attachment of 12. System on Chip/SoCs according to claim 11, the method also comprises step:
With described first transfer rate, the enable signal of described service load and described first clock signal synchronization are sent to described second interface from described first interface by described first channel.
The bus interface method of attachment of 13. System on Chip/SoCs according to claim 12, the method also comprises step:
With the second transfer rate, second clock signal and the answer signal synchronous with described second clock signal are sent to described first interface from described second interface by the 3rd channel.
The bus interface method of attachment of 14. System on Chip/SoCs according to claim 13, wherein, performs bus interface based on advanced extensible interface bus agreement and connects.
The bus interface method of attachment of 15. System on Chip/SoCs according to claim 14, wherein, when the frequency of described first clock signal is equal to or greater than 500Mhz, the length of described first channel is greater than 2000 μm.
16. 1 kinds of bus interface connecting circuit, comprising:
Sender interface, it is configured to send service load, write enable signal and transmitter clock signal by the first channel; And
Receiver interface, it comprises:
Push-up storage, its remote write pointer be configured to based on being produced based on described write enable signal by described receiver interface stores described service load; And
Service load receiver, it is configured to read described service load from described push-up storage,
Wherein, described receiver interface is by second channel transmitter-receiver clock and answer signal, and the length of described second channel corresponds to the length of described first channel.
17. bus interface connecting circuit according to claim 16, wherein, are sent to described sender interface by service load from memory assembly.
18. bus interface connecting circuit according to claim 17, wherein, described service load receiver is connected to Memory Controller, and described Memory Controller is configured to control described memory assembly.
19. bus interface connecting circuit according to claim 18, wherein, synchronously latch described service load and described transmitter clock signal and are sent to described receiver interface.
20. bus interface connecting circuit according to claim 19, wherein, are synchronously stored in described service load and described transmitter clock signal in described push-up storage.
CN201510086595.1A 2014-02-20 2015-02-17 System chip, bus interface connection circuit and bus interface connection method thereof Active CN104866453B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911149385.7A CN111008170B (en) 2014-02-20 2015-02-17 System chip, bus interface connection circuit and bus interface connection method thereof

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
KR10-2014-0019742 2014-02-20
KR20140019742 2014-02-20
US14/486,434 US9489009B2 (en) 2014-02-20 2014-09-15 System on chip, bus interface and method of operating the same
US14/486,434 2014-09-15

Related Child Applications (1)

Application Number Title Priority Date Filing Date
CN201911149385.7A Division CN111008170B (en) 2014-02-20 2015-02-17 System chip, bus interface connection circuit and bus interface connection method thereof

Publications (2)

Publication Number Publication Date
CN104866453A true CN104866453A (en) 2015-08-26
CN104866453B CN104866453B (en) 2019-12-13

Family

ID=53799110

Family Applications (2)

Application Number Title Priority Date Filing Date
CN201911149385.7A Active CN111008170B (en) 2014-02-20 2015-02-17 System chip, bus interface connection circuit and bus interface connection method thereof
CN201510086595.1A Active CN104866453B (en) 2014-02-20 2015-02-17 System chip, bus interface connection circuit and bus interface connection method thereof

Family Applications Before (1)

Application Number Title Priority Date Filing Date
CN201911149385.7A Active CN111008170B (en) 2014-02-20 2015-02-17 System chip, bus interface connection circuit and bus interface connection method thereof

Country Status (5)

Country Link
US (2) US9489009B2 (en)
JP (1) JP2015156645A (en)
KR (1) KR102336452B1 (en)
CN (2) CN111008170B (en)
TW (1) TWI695269B (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109597785A (en) * 2018-12-04 2019-04-09 中国航空工业集团公司西安航空计算技术研究所 A kind of design method of the PCI Express bus adaption multiplex roles based on AXI4 interface
WO2019144267A1 (en) * 2018-01-23 2019-08-01 深圳市大疆创新科技有限公司 Chip, processor, computer system, and mobile device
CN110691021A (en) * 2015-09-03 2020-01-14 阿尔特拉公司 Distributed multi-chip protocol application interface
CN112306924A (en) * 2020-11-26 2021-02-02 山东云海国创云计算装备产业创新中心有限公司 Data interaction method, device and system and readable storage medium
CN112667533A (en) * 2019-10-16 2021-04-16 三星电子株式会社 Semiconductor device including subsystem interface and communication method thereof
CN113474761A (en) * 2019-02-27 2021-10-01 微芯片技术股份有限公司 Programmable arbitrary sequence direct memory access controller for configuring multiple core independent peripherals

Families Citing this family (24)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP5935235B2 (en) * 2011-02-18 2016-06-15 ソニー株式会社 COMMUNICATION DEVICE, COMMUNICATION SYSTEM, AND COMMUNICATION METHOD
WO2012158392A2 (en) * 2011-05-17 2012-11-22 Rambus Inc. Memory system using asymmetric source-synchronous clocking
US9489009B2 (en) 2014-02-20 2016-11-08 Samsung Electronics Co., Ltd. System on chip, bus interface and method of operating the same
US10013375B2 (en) 2014-08-04 2018-07-03 Samsung Electronics Co., Ltd. System-on-chip including asynchronous interface and driving method thereof
DE102016109387A1 (en) 2015-05-26 2016-12-01 Samsung Electronics Co., Ltd. One-chip system with clock management unit and method for operating the one-chip system
KR102384347B1 (en) * 2015-05-26 2022-04-07 삼성전자주식회사 System on chip includiing clock management unit and operating method thereof
KR102571154B1 (en) * 2016-02-03 2023-08-24 삼성전자주식회사 Semiconductor device, semiconductor system and method for operating semiconductor device
KR102497804B1 (en) * 2016-04-01 2023-02-10 한국전자통신연구원 On-chip network device capable of networking in dual swithching network modes and operation method thereof
US10168923B2 (en) 2016-04-26 2019-01-01 International Business Machines Corporation Coherency management for volatile and non-volatile memory in a through-silicon via (TSV) module
CN106383573B (en) * 2016-11-09 2023-04-28 杭州电子科技大学 Low-power-consumption self-return type releaser circuit
US10649929B2 (en) * 2017-07-10 2020-05-12 Dialog Semiconductar Korea Inc. Memory time-sharing method and apparatus capable of distributing bus traffic of system-on-chip
KR102607349B1 (en) * 2019-01-15 2023-11-29 한국전자통신연구원 Network on chip using temperature effect inversion
WO2020172557A1 (en) 2019-02-22 2020-08-27 Micron Technology, Inc. Memory device interface and method
CN110569173B (en) * 2019-09-16 2022-12-27 超越科技股份有限公司 Server health management chip based on Loongson IP core and implementation method
US11989141B2 (en) 2019-12-27 2024-05-21 Micron Technology, Inc. Neuromorphic memory device and method
KR20220116258A (en) 2019-12-30 2022-08-22 마이크론 테크놀로지, 인크. Memory device interfaces and methods
US11538508B2 (en) * 2019-12-31 2022-12-27 Micron Technology, Inc. Memory module multiple port buffer techniques
TWI764139B (en) * 2020-04-27 2022-05-11 鴻海精密工業股份有限公司 Data bus accessing device, method and system
WO2022052918A1 (en) * 2020-09-11 2022-03-17 北京灵汐科技有限公司 Signal transmission method and apparatus
CN117561489A (en) * 2021-06-22 2024-02-13 谷歌有限责任公司 Independent clocking of configuration and status registers
CN113611238B (en) * 2021-07-26 2023-03-14 珠海海奇半导体有限公司 Display control circuit and display system based on 8080 bus interface
KR20230016816A (en) * 2021-07-27 2023-02-03 주식회사 세미파이브 Interface method for transmitting and receiving data between functional blocks in system-on-chip and system-on-chip using the same
CN116701297B (en) * 2023-08-08 2024-02-27 深圳比特微电子科技有限公司 Synchronous bridge circuit, bus device and system on chip
CN116909975B (en) * 2023-09-12 2024-01-26 苏州浪潮智能科技有限公司 Multi-master multi-slave interaction control system of serial bus standard

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7286487B2 (en) * 2002-11-18 2007-10-23 Infinera Corporation Optical transmission network with asynchronous mapping and demapping and digital wrapper frame for the same
US7366803B1 (en) * 2005-02-23 2008-04-29 Xilinx, Inc. Integrated circuit for buffering data by removing idle blocks to create a modified data stream when memory device is not near empty
US7793021B2 (en) * 2006-01-05 2010-09-07 Freescale Semiconductor, Inc. Method for synchronizing a transmission of information and a device having synchronizing capabilities

Family Cites Families (45)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
IT1161467B (en) 1983-01-21 1987-03-18 Cselt Centro Studi Lab Telecom PARALLEL INTERFACE FOR INTERVIEW MANAGEMENT BETWEEN AN ASYNCHRONOUS BUS AND A SYNCHRONOUS BUS CONNECTED TO MULTIPLE TERMINALS EQUIPPED EACH WITH ITS OWN SYNCHRONIZATION SIGNAL
US4569062A (en) * 1984-06-28 1986-02-04 Dellande Brian W Interface circuit for interfacing between asynchronous data in start/stop format and synchronous data
DE3751608T2 (en) 1986-09-01 1996-06-27 Nippon Electric Co Serial bus interface system for data transmission with a two-wire line as clock bus and data bus.
US5199106A (en) 1986-09-19 1993-03-30 International Business Machines Corporation Input output interface controller connecting a synchronous bus to an asynchronous bus and methods for performing operations on the bus
EP0606102A1 (en) 1986-09-19 1994-07-13 International Business Machines Corporation An input output interface controller connecting a synchronous bus to an asynchronous bus and methods for performing operations on the buses
US5079696A (en) 1989-09-11 1992-01-07 Sun Microsystems, Inc. Apparatus for read handshake in high-speed asynchronous bus interface
US5070443A (en) 1989-09-11 1991-12-03 Sun Microsystems, Inc. Apparatus for write handshake in high-speed asynchronous bus interface
GB9019001D0 (en) * 1990-08-31 1990-10-17 Ncr Co Work station including a direct memory access controller and interfacing means to microchannel means
US5265216A (en) 1991-06-28 1993-11-23 Digital Equipment Corporation High performance asynchronous bus interface
JPH05160825A (en) * 1991-12-03 1993-06-25 Fujitsu Ltd Synchronizing circuit
US5255375A (en) 1992-01-10 1993-10-19 Digital Equipment Corporation High performance interface between an asynchronous bus and one or more processors or the like
US5469547A (en) 1992-07-17 1995-11-21 Digital Equipment Corporation Asynchronous bus interface for generating individual handshake signal for each data transfer based on associated propagation delay within a transaction
US5592685A (en) 1992-10-07 1997-01-07 Digital Equipment Corporation Synchronous/asynchronous partitioning of an asynchronous bus interface
FI104858B (en) 1995-05-29 2000-04-14 Nokia Networks Oy Method and apparatus for adapting an asynchronous bus to a synchronous circuit
KR0157924B1 (en) 1995-12-23 1998-12-15 문정환 Data transfer system and its method
JP3733699B2 (en) 1997-06-20 2006-01-11 ソニー株式会社 Serial interface circuit
US6167062A (en) * 1998-02-02 2000-12-26 Tellabs Operations, Inc. System and associated method for the synchronization and control of multiplexed payloads over a telecommunications network
JP3711730B2 (en) 1998-02-06 2005-11-02 富士ゼロックス株式会社 Interface circuit
US6345330B2 (en) * 1998-05-01 2002-02-05 Acqis Technology, Inc. Communication channel and interface devices for bridging computer interface buses
US6418494B1 (en) 1998-10-30 2002-07-09 Cybex Computer Products Corporation Split computer architecture to separate user and processor while retaining original user interface
JP2001014269A (en) 1999-06-29 2001-01-19 Toshiba Corp Computer system
US6850092B2 (en) 2000-06-09 2005-02-01 The Trustees Of Columbia University Low latency FIFO circuits for mixed asynchronous and synchronous systems
JP3523181B2 (en) 2000-11-01 2004-04-26 Necマイクロシステム株式会社 Asynchronous bus interface circuit and microcomputer
KR100357284B1 (en) 2001-02-19 2002-10-18 (주)씨앤에스 테크놀로지 The one chip asynchronous microprocessor Inter Processor Communication circuit
DE10119472A1 (en) 2001-04-20 2002-10-31 Harman Becker Automotive Sys Interface for data transmission between two bus systems and operating procedures therefor
US6795882B1 (en) 2001-06-04 2004-09-21 Advanced Micro Devices, Inc. High speed asynchronous bus for an integrated circuit
JP3853188B2 (en) 2001-10-04 2006-12-06 Necエレクトロニクス株式会社 Asynchronous bus interface device
US7120327B2 (en) 2002-11-27 2006-10-10 International Business Machines Corporation Backplane assembly with board to board optical interconnections
US7408878B2 (en) * 2003-06-10 2008-08-05 Cisco Technology, Inc. System packet interface
US7467358B2 (en) 2004-06-03 2008-12-16 Gwangju Institute Of Science And Technology Asynchronous switch based on butterfly fat-tree for network on chip application
US7184360B2 (en) * 2005-06-15 2007-02-27 Infineon Technologies, Ag High-speed interface circuit for semiconductor memory chips and memory system including semiconductor memory chips
JP2007048022A (en) 2005-08-10 2007-02-22 Fujitsu Ltd Asynchronous bus interface and its processing method
US7783820B2 (en) 2005-12-30 2010-08-24 Avocent Corporation Packet-switched split computer having disassociated peripheral controller and plural data buses
AU2007215381B2 (en) * 2006-02-15 2012-06-28 Chronologic Pty. Ltd. Distributed synchronization and timing system
US7657668B2 (en) 2006-08-16 2010-02-02 Qnx Software Systems (Wavemakers), Inc. Clock synchronization of data streams
US7590788B2 (en) 2007-10-29 2009-09-15 Intel Corporation Controlling transmission on an asynchronous bus
JP2011065415A (en) 2009-09-17 2011-03-31 Yaskawa Electric Corp Asynchronous bus interface circuit
JP5340107B2 (en) 2009-10-13 2013-11-13 日本圧着端子製造株式会社 Coaxial connector and coaxial connector with antenna
KR20110058575A (en) * 2009-11-26 2011-06-01 삼성전자주식회사 Bandwidth synchronization circuit and thereof method in data processing system
KR20110061189A (en) * 2009-12-01 2011-06-09 삼성전자주식회사 Asynchronization upsizing circuit in data processing system
JP5560833B2 (en) * 2010-03-29 2014-07-30 富士通株式会社 Optical interface device and input frequency deviation abnormality monitoring method
JP5617429B2 (en) * 2010-08-19 2014-11-05 ソニー株式会社 Bridge system for connecting the bus system and the bus system to the connected device
US8793095B2 (en) * 2011-03-09 2014-07-29 Intel Corporation Functional fabric-based test controller for functional and structural test and debug
CN102508808B (en) * 2011-11-14 2014-07-16 北京北大众志微***科技有限责任公司 System and method for realizing communication of master chip and extended chip
US9489009B2 (en) 2014-02-20 2016-11-08 Samsung Electronics Co., Ltd. System on chip, bus interface and method of operating the same

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7286487B2 (en) * 2002-11-18 2007-10-23 Infinera Corporation Optical transmission network with asynchronous mapping and demapping and digital wrapper frame for the same
US7366803B1 (en) * 2005-02-23 2008-04-29 Xilinx, Inc. Integrated circuit for buffering data by removing idle blocks to create a modified data stream when memory device is not near empty
US7793021B2 (en) * 2006-01-05 2010-09-07 Freescale Semiconductor, Inc. Method for synchronizing a transmission of information and a device having synchronizing capabilities

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110691021A (en) * 2015-09-03 2020-01-14 阿尔特拉公司 Distributed multi-chip protocol application interface
CN110691021B (en) * 2015-09-03 2022-02-15 阿尔特拉公司 Distributed multi-chip protocol application interface
WO2019144267A1 (en) * 2018-01-23 2019-08-01 深圳市大疆创新科技有限公司 Chip, processor, computer system, and mobile device
CN109597785A (en) * 2018-12-04 2019-04-09 中国航空工业集团公司西安航空计算技术研究所 A kind of design method of the PCI Express bus adaption multiplex roles based on AXI4 interface
CN113474761A (en) * 2019-02-27 2021-10-01 微芯片技术股份有限公司 Programmable arbitrary sequence direct memory access controller for configuring multiple core independent peripherals
CN112667533A (en) * 2019-10-16 2021-04-16 三星电子株式会社 Semiconductor device including subsystem interface and communication method thereof
CN112667533B (en) * 2019-10-16 2024-05-28 三星电子株式会社 Semiconductor device including subsystem interface and communication method thereof
CN112306924A (en) * 2020-11-26 2021-02-02 山东云海国创云计算装备产业创新中心有限公司 Data interaction method, device and system and readable storage medium

Also Published As

Publication number Publication date
CN111008170B (en) 2023-11-24
TWI695269B (en) 2020-06-01
JP2015156645A (en) 2015-08-27
TW201533577A (en) 2015-09-01
KR20150099394A (en) 2015-08-31
US9811482B2 (en) 2017-11-07
US20170052910A1 (en) 2017-02-23
CN111008170A (en) 2020-04-14
US20150236870A1 (en) 2015-08-20
CN104866453B (en) 2019-12-13
US9489009B2 (en) 2016-11-08
KR102336452B1 (en) 2021-12-08

Similar Documents

Publication Publication Date Title
CN104866453A (en) System on a chip, bus interface connection circuit and method for connecting a bus interface
CN109478177A (en) Double data rate (DDR) command line
CN105027444B (en) Integrated clock differential buffer
KR20140067404A (en) Storage device and computing system havint its and data transfering method thereof
KR101733203B1 (en) Bus system having id converter and coverting method thereof
US10303366B2 (en) Data storage device that divides and processes a command and data processing system including the same
CN108536642A (en) Big data operation acceleration system and chip
Markettos et al. Interconnect for commodity FPGA clusters: Standardized or customized?
US9632953B2 (en) Providing input/output virtualization (IOV) by mapping transfer requests to shared transfer requests lists by IOV host controllers
CN108091355A (en) Data antiphase circuit
US11789885B2 (en) Ordered delivery of data packets based on type of path information in each packet
KR102101840B1 (en) Universal serial interface and semiconductor device comprising the same
CN101436171A (en) Modular communication control system
US20140173228A1 (en) Memory system and system on chip including the same
US11714655B2 (en) Pipeline merging in a circuit
Khalifa et al. Memory controller architectures: A comparative study
Therdsteerasukdi et al. Utilizing radio-frequency interconnect for a many-DIMM DRAM system
CN208298179U (en) Big data operation acceleration system and chip
Hsiao et al. Asynchronous AHB bus interface designs in a multiple-clock-domain graphics system
CN103226537A (en) Programmable logic device for implementing hardware interface of mobile phone
US11488643B2 (en) Method for configuring multiple input-output channels
US11455262B2 (en) Reducing latency for memory operations in a memory controller
TWI810815B (en) Controller and memory system
US8347258B2 (en) Method and apparatus for interfacing multiple dies with mapping for source identifier allocation
WO2022115167A2 (en) Multiple channel memory system

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant